EP2031530A2 - Procédé d'analyse assistée par ordinateur d'un objet - Google Patents

Procédé d'analyse assistée par ordinateur d'un objet Download PDF

Info

Publication number
EP2031530A2
EP2031530A2 EP08104955A EP08104955A EP2031530A2 EP 2031530 A2 EP2031530 A2 EP 2031530A2 EP 08104955 A EP08104955 A EP 08104955A EP 08104955 A EP08104955 A EP 08104955A EP 2031530 A2 EP2031530 A2 EP 2031530A2
Authority
EP
European Patent Office
Prior art keywords
time
description
reasoners
analysis
behavior
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Withdrawn
Application number
EP08104955A
Other languages
German (de)
English (en)
Other versions
EP2031530A3 (fr
Inventor
Michael Berger
Dagmar Beyer
Thomas Hubauer
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Siemens AG
Original Assignee
Siemens AG
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Siemens AG filed Critical Siemens AG
Publication of EP2031530A2 publication Critical patent/EP2031530A2/fr
Publication of EP2031530A3 publication Critical patent/EP2031530A3/fr
Withdrawn legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H40/00—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
    • G16H40/60—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices
    • G16H40/67—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices for remote operation

Definitions

  • the invention relates to a method for computer-aided analysis of an object and to a corresponding computer program product.
  • Behavioral recognition plays an essential role in a large number of technical areas.
  • the condition of a patient can be determined via behavioral recognition and acute emergencies can be detected.
  • security technology for example, a person's behavior can be interpreted as a break-in attempt.
  • traffic engineering recognizing the behavior of the driver and the vehicle plays an important role in driver assistance systems.
  • a first class Such methods are limited to viewing the system and classifying the situation at a given time. The lack of observation of the temporal change of the situation does not make it possible to derive a corresponding behavior.
  • the object of the invention is to provide a method for computer-aided analysis of an object, which simply and efficiently recognizes the temporal behavior of the object.
  • object in each case information about the object with one or more sensor means is detected for a plurality of temporally successive times and / or periods.
  • object is here to be interpreted broadly and can affect both objects and persons.
  • An object may include, for example, a living being to be monitored, in particular a human whose behavior is analyzed, for example, in a medical monitoring system.
  • an object may comprise a spatial area with objects and / or persons contained therein, for example a room or a room.
  • the object may also relate to any technical systems or parts thereof.
  • the object can be an automation system and / or a part of an automation system and / or a transport system and / or a part of a transport system, for example a rail vehicle.
  • the object may further comprise a vehicle and the driver of a vehicle whose behavior is analyzed in a driver assistance system.
  • object is also to be understood as meaning that an object may comprise many different components and is not limited to a single object or person.
  • a first analysis method is applied to the information acquired for the respective time and / or period, whereby state descriptions of the object at the respective times and / or Periods are obtained.
  • a plurality of state descriptions of the object is then applied to a further, second analysis method, whereby a temporal behavioral description of the object is obtained.
  • This second analysis method can also be repeatedly applied to newly acquired state descriptions to detect changes in behavior.
  • the state descriptions at the various times thus again represent inputs to the second analysis method, i. they are perceived as sensory information.
  • the first or second analysis method used is an ontology-based inference method.
  • Ontology defines the knowledge about the object by terms and relations, and with the help of the inference method, the corresponding state or behavior descriptions are derived from the ontology.
  • Ontology-based inference methods are well known in the computer science field.
  • inference methods are used which are based on a deterministic or probabilistic description logic, wherein the state description or behavioral description of the object is determined with the aid of one or more reasoners.
  • OWL-DL Web Ontology Language Description Language
  • a plurality of reasoners are used for carrying out the first analysis method, the reasoners being applied to information on the object which was acquired at different, in particular successive, times and / or periods.
  • a particularly simple combination of a state description of the object at a respective time and / or period with the information about the time and / or period is characterized achieved that the respective state description is assigned a corresponding timestamp.
  • the first or second analysis step may optionally also derive a corresponding state or behavioral description of the object in another way.
  • causal networks well known in the art may be used for this, e.g. Bayesian networks.
  • the behavioral description determined using the second analysis method is compared with predetermined behavioral descriptions, and in the event of deviation from the predetermined behavioral descriptions, a message, in particular an alarm signal, is output.
  • a message in particular an alarm signal
  • the sensor means with which information about the object are detected, can be configured as desired. This may be, for example, a camera and / or a motion sensor and / or a sound sensor and / or sensors for recording vital signs of a living being.
  • the invention further relates to a computer program product with a program code stored on a machine-readable carrier for carrying out the method according to the invention when the program runs on a computer.
  • a situation at these points in time relating to an object to be analyzed O is detected.
  • the situation identification is designated in FIG. 1 by way of example for the time steps t1, t2 or tn with SR1, SR2 or SRn.
  • the correspondingly recognized situation is reproduced for the time steps t1, t2 and tn with S1, S2 and Sn.
  • the individual situations are identified by means of corresponding sensor means which detect the object to be analyzed.
  • the object can in this case relate to any observable scenario.
  • the invention could be used in a medical monitoring system where the observed object is a human person.
  • the sensor means may relate, for example, to sensors which detect whether light is switched on in the room or not.
  • the sensors may relate to noise sensors, with which it can be determined whether speech is being made in the room.
  • additional sensors may be integrated into the chairs of the meeting room to determine if and how many people are present in the meeting room. It could also be determined whether a person is standing at the desk in the meeting room, for example, with a corresponding camera whose image is analyzed. From all these variables, a corresponding situation can then be identified at a certain time, for example, whether the room is free, whether a lecture is being held or whether a discussion is taking place.
  • a situation detection relates to the detection of parked luggage, for example in the waiting room of a train station.
  • the events in the waiting room can be followed by means of a video camera and by means of an image analysis of the recorded video material, persons and objects as well as the spatial relationships between them can be recognized as a situation.
  • An example of such a recognized situation is the state description "Person P1 carries item O2". The objects and their relationships thus correspond to the individual situations S1, S2,..., Sn according to FIG. 1.
  • a situation can be detected in any manner at successive times or periods of any objects.
  • computer-assisted analysis methods based on so-called reasoners are used for situation recognition.
  • Reasoners provide methods for deriving conclusions based on a representation of knowledge, thereby allowing new knowledge to be determined.
  • Reasoners usually use a description logic, which is a subset of the predicate logic known to those skilled in the art. The description logic is characterized by the fact that it is decidable. Description logics usually subdivide the represented knowledge into one so-called T-box, which describes all concepts of the model shown, as well as an A-box, which contains all instances of the concepts, as well as their relationships to each other.
  • a concept describes a class of similar objects, whereas an instance represents a concrete object.
  • one concept could be the generic term "suitcase”, whereas one instance of the concept stands for a single, uniquely identifiable suitcase.
  • Description logics are often used to represent an ontology that represents the knowledge representation of a formally defined system of terms and relations. Description logics as well as ontologies are well known in the art, and therefore will not be discussed in detail here.
  • the ontology language OWL-DL is used to describe the ontology underlying the object under consideration. Based on this language for knowledge representation conclusions are then derived with a corresponding reasoner. Reasoners are also well known in the art and therefore will not be discussed in detail. According to the embodiment of the invention described here, racers, pellets or FaCT ++ can be used as reasonants, for example.
  • corresponding situations S1, S2,..., Sn are first separately detected in respective time steps t1, t2,..., Tn, for example based on description logics with the aid of reasoners.
  • Time information which can be explicitly given by time stamps, the use of temporal logics and the like, but which can also be implicitly given in the individual recognized situations, for example by a successor relation, are assigned to the individual situations.
  • situation changes occur between the individual time steps t1, t2, etc., the situation change SC1 between the time steps t1 and t2 being reproduced in FIG. 1, and a situation change SC2 representing a plurality of time steps between t2 and tn includes.
  • the situation changes here have their cause in a corresponding change in the behavior of the observed object, wherein the behavior change associated with the situation change SC1 is denoted by BC1 and the behavior change associated with the situation SC2 is designated by BC2.
  • An essential aspect of the method according to the invention is now to recognize the corresponding change in behavior based on the changing situations and to derive a temporal behavioral description. This is done with a further analysis method, whereby this analysis method is no longer used directly on quantities detected with sensor means.
  • the individual detected situations S1, S2,..., Sn represent the input variables of the further analysis method.
  • This further analysis method is designated BR1 or BR2 in FIG. 1, wherein the method BR1 is a behavioral description from the situation change SC1 and the method BR2 derives a behavioral description from the situation change SC2.
  • a reasoner based on a description logic can again be used here, which now receives situations at successive times as input variables and, based thereon on the description logic, derives new knowledge, which is represented by a behavioral description of the object.
  • reasoners are used for situation recognition in the individual time steps t1, t2,..., Tn, and a further reasoner is then used to derive a behavioral description from the temporally successive situations.
  • a further reasoner is then used to derive a behavioral description from the temporally successive situations.
  • several reasoners are used parallel to the situation recognition for several points in time.
  • situations can be detected at close intervals since there is no need to wait until the reasoner has completed the situation recognition and can perform situation recognition for a new time. Rather, a new situation can already be analyzed with a reasoner used in parallel, while the reasoner used at a previous time is still busy analyzing the situation.
  • any objects can be analyzed with the method according to the invention.
  • deviations from previous behavioral patterns can be recognized from the individual identified situations in the form of behavioral profiles, and from this e.g. early symptoms of depression are detected.
  • the depressive behavior can result from corresponding deviations of the behavior of the person with regard to sleep times and activity times. For example, in the above example regarding the detection of situations in a meeting room, it could be determined when a meeting in the room starts or ends.
  • suspected behaviors may be deduced from the individual situations concerning a baggage and a person who has parked the baggage.
  • the suspicious behavior can be recognized that a person has parked a suitcase and subsequently removed himself from the object.
  • This behavior in particular indicates that there is a dangerous object in the suitcase, for example explosive material.
  • the inventive method could thus be used in a monitoring system in a train station, wherein when a suspicious behavior a warning message is issued to a control center, whereupon appropriate security measures, such as the evacuation of the station, can be initiated.
  • Another field of application of the invention is the monitoring of an automation system or a part of an automation system, wherein the state of the automation system is in turn detected with corresponding sensor means at predetermined times and from this a behavioral description of the automation system is derived, for example, to detect suspicious events in the automation system and to initiate the temporary shutdown of the system to prevent further damage.
  • the method provides a generic time-based behavior detection system configurable by exchanging the underlying ontology.
  • the temporal component By explicitly considering the temporal component, more accurate statements can be made than would be possible by simply looking at "snapshots" at given times.
  • description logic reasoners for situational and behavioral recognition allows a compromise between expressiveness and decidability as well as the use of proven and efficient building blocks.
  • probabilistic description logics can be used in the invention, which prioritize situations and behaviors according to their "relevance" and thus provide significant added value, especially in the field of decision support.
EP08104955A 2007-08-09 2008-08-04 Procédé d'analyse assistée par ordinateur d'un objet Withdrawn EP2031530A3 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
DE102007037633 2007-08-09
DE102007054835A DE102007054835A1 (de) 2007-08-09 2007-11-16 Verfahren zur rechnergestützten Analyse eines Objekts

Publications (2)

Publication Number Publication Date
EP2031530A2 true EP2031530A2 (fr) 2009-03-04
EP2031530A3 EP2031530A3 (fr) 2010-03-10

Family

ID=40227071

Family Applications (1)

Application Number Title Priority Date Filing Date
EP08104955A Withdrawn EP2031530A3 (fr) 2007-08-09 2008-08-04 Procédé d'analyse assistée par ordinateur d'un objet

Country Status (2)

Country Link
EP (1) EP2031530A3 (fr)
DE (1) DE102007054835A1 (fr)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE102011075231A1 (de) * 2011-05-04 2012-11-08 Siemens Ag Verfahren und Vorrichtung zum Betrieb einer Anlage
DE102016100968A1 (de) 2016-01-21 2017-07-27 Dr. Ing. H.C. F. Porsche Aktiengesellschaft Verfahren und Vorrichtung zum Überwachen eines Fahrzeugführers

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
AU2003245157A1 (en) * 2002-06-20 2004-01-06 Andromed Inc. Telehealth system and method
US8589174B2 (en) * 2003-12-16 2013-11-19 Adventium Enterprises Activity monitoring
US7944468B2 (en) * 2005-07-05 2011-05-17 Northrop Grumman Systems Corporation Automated asymmetric threat detection using backward tracking and behavioral analysis

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
HOW-LUNG ENG ET AL.: "An automatic drowning detection surveillance system for challenging outdoor pool environments", PROC. OF THE NINTH IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2003

Also Published As

Publication number Publication date
EP2031530A3 (fr) 2010-03-10
DE102007054835A1 (de) 2009-02-12

Similar Documents

Publication Publication Date Title
DE112016005965B4 (de) Vorrichtung zur überwachung unbeaufsichtigter objekte, mit derselben versehenes system zur überwachung unbeaufsichtigter objekte und verfahren zur überwachung unbeaufsichtigter objekte
EP0986912B1 (fr) Procede pour surveiller une zone de surveillance predefinie
DE602005005583T2 (de) Videoüberwachungssysteme im allgemeinen und speziell eine Vorrichtung und Verfahren zur automatischen Echtzeiterfassung abnormaler Bewegung in Videoströmen
DE602004012195T2 (de) Integriertes elektronisches Artikelüberwachungsetikett und -system zum Zählen von Personen
EP1346330B1 (fr) Systeme de detection de fumee base sur la video
EP3499473A1 (fr) Détection automatisée de situations dangereuses
EP2462534B1 (fr) Procédé de surveillance d'une zone
AT17967U1 (de) Alarmsituation
EP2031530A2 (fr) Procédé d'analyse assistée par ordinateur d'un objet
EP2483834B1 (fr) Methode et appareil pour la reconnaissance d'une detection fausse d'un objet dans un image
EP1227425B1 (fr) Système pour découvrir la source des attaques de maladies
DE102019007319A1 (de) Kamerabasiertes Wasserbereich Überwachungssystem
WO2012110654A1 (fr) Procédé pour analyser une pluralité d'images décalées dans le temps, dispositif pour analyser des images, système de contrôle
WO2013083327A1 (fr) Dispositif et procédé de détection automatique d'événements dans des données de capteur
DE102017219269A1 (de) Klassifizierung mit automatischer Auswahl aussichtsreicher Lerndaten
EP3542351A1 (fr) Dispositif de surveillance pour surveiller une zone de surveillance et système de surveillance équipé du dispositif de surveillance
WO2007028698A1 (fr) Caméra de surveillance intelligente
DE19641000A1 (de) Verfahren und Anordnung zur automatischen Erkennung der Anzahl von Personen in einer Personenschleuse
DE102008000836A1 (de) Verfahren zum Überwachen
DE102019204269A1 (de) Verfahren und Einrichtung zum Erzeugen eines Statussignals
DE102022200833A1 (de) Überwachungsanordnung, Verfahren zur Registrierung von Überwachungskameras und Analysemodulen, Computerprogramm sowie Speichermedium
AT502241B1 (de) Verfahren und anordnung zur feststellung der abweichung von ermittelten werten
WO2017167656A1 (fr) Procédé de surveillance à distance d'une installation industrielle
DE102022114494A1 (de) Kollaborative Robotervorrichtung mit Sensorik zur Berücksichtigung des menschlichen Wohlbefindens in einer geteilten Arbeitsumgebung
DE102022207966A1 (de) Verfahren, Recheneinheit und Computerprogramm

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

AK Designated contracting states

Kind code of ref document: A2

Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MT NL NO PL PT RO SE SI SK TR

AX Request for extension of the european patent

Extension state: AL BA MK RS

PUAL Search report despatched

Free format text: ORIGINAL CODE: 0009013

AK Designated contracting states

Kind code of ref document: A3

Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MT NL NO PL PT RO SE SI SK TR

AX Request for extension of the european patent

Extension state: AL BA MK RS

AKY No designation fees paid
STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: THE APPLICATION IS DEEMED TO BE WITHDRAWN

18D Application deemed to be withdrawn

Effective date: 20100911

REG Reference to a national code

Ref country code: DE

Ref legal event code: R108

Effective date: 20110309

Ref country code: DE

Ref legal event code: 8566